Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

GPT-4 details leaked?

threadreaderapp.com

Post reply on HN
Print view View original

91–100 of 648 posts

Re: GPT-4 details leaked?

#91

Post by EagnaIonat »

PostOnce wrote: post #38
BoorishBears wrote: post #34

Earlier quoted context omitted.

Why the vitriol towards OpenAI? If Elon hadn't pulled the rug out from under them after they refused his forceful takeover*, they wouldn't have had to go to Microsoft and they'd still be open. * a takeover which he predicated on the claim that OpenAI was "doomed to fail"

It's several reasons, first it's the lies and the abuse of a charity, that wouldn't be an issue if they had began as a private company instead of robbing a charity. But secondly, even if they were a private company, it's dishonest and reprehensible to claim to congress that you want to "protect the public" when you really only give a shit about protecting your moat, I'm not happy about that either. I'm also tired of…

> the abuse of a charity,

A non-profit and charity are two different things.

While a charity is a form of non-profit, it has to follow certain rules to qualify as one. Their profits must go towards the charity.

A non-profit is a company that is set up not to make a profit. It is allowed to make a profit if it does. This is what OpenAI was.

They switched to a "capped" for-profit model so that they could get more funding. It also allowed their employees to invest in the company, and openAI gave equity to their employees.

There was no lies or abuse. Where did you get that information from?

Re: GPT-4 details leaked?

#92

Post by antupis »

KronisLV wrote: post #76
PostOnce wrote: post #19

"Open" AI, a charity to benefit us all by pushing and publishing the frontier of scientific knowledge. Nevermind, fuckers, actually it's just to take your jobs and make a few VCs richer. We'll keep the science a secret and try to pressure the government into making it illegal for you to compete with us. https://github.com/ggerganov/llama.cpp https://github.com/openlm-research/open_llama https://huggingface.co/TheBlok…

> You can use the above without paying OpenAI. You don't even need a GPU. There are no license issues like with the facebook llama. I actually wrote about getting an LLM chatbot up and running a while ago: https://blog.kronis.dev/tutorials/self-hosting-an-ai-llm-cha... It's good that the technology and models are both available for free, and you don't even need a GPU for it. However, there are still large memory requ…

I bet that open models win in the end because porn. There is already very weird and vibrant community creating "waifus" and tinkering with these models.

Re: GPT-4 details leaked?

#93

Post by swyx »

refulgentis wrote: post #31
jph00 wrote: post #8

Earlier quoted context omitted.

No, a meme post from 2022 did not in fact reference papers posted in 2023. You must be thinking of some other post, or you're just making stuff up.

Like I said, it gains a little bit in every retelling. Why are you aggressively defending unsourced tripe?

well as someone out of the loop - is there a source on the Q4 2022 version then?

Re: GPT-4 details leaked?

#94

Post by Ozzie_osman »

There's a section at the end where there is speculation on what the entire dataset entails. My guess is a chunk of it is probably from ChatGPT data (or GPT3 data from when training on your requests was opt-out rather than opt-in).

Re: GPT-4 details leaked?

#95

Post by vouaobrasil »

M3L0NM4N wrote: post #87
vouaobrasil wrote: post #80

Earlier quoted context omitted.

Not sure if you're being sarcastic but I agree. Humans should not have AI research or advanced AI at all. It (a) removes purpose from people, (b) presents a situation that is too alien for human minds to handle, (c) increases the addictiveness of technology and thereby pushes us further into growing the technological system, (d) crosses the "adaptability threshhold", i.e. the point at which the PACE of technological…

This is such a pessimistic view to take that I cannot even begin to describe how you are wrong.

I would love to see your rebuttals, especially since I have never seen any strong arguments in favour of AI being a net benefit to society, and I have thought and read about this at lenght.

Of course, I always expect downvotes on my posts here since there is a strong tendency towards loving technology here. But what I find most interesting is that there is absolutely no taking of responsibility of any technological creations.

Before you downvote, please just ask yourself the following: is it reasonable to say that all these latest AI developments have no serious risks? And in response to your reply, humanity is in quite a precarious state now, so isn't it expected that a sober analysis of it is rather pessimistic, especially with regard to hyper-advanced technologies?

Re: GPT-4 details leaked?

#96

Post by mmahemoff »

I've been wondering how freemium services like Thread Reader still operate now that Twitter is charging prohibitive prices for API access and taking measures to prevent scraping. The cheapest API plan with read access is $100/month, which reads 10,000 tweets, so could only produce about 500 pages like this one on demand.

Re: GPT-4 details leaked?

#97

Post by schappim »

Roark66 wrote: post #79
kyledrake wrote: post #65

Earlier quoted context omitted.

Unfortunately I've found the current OSS models to be vastly inferior to the OpenAI models. Would love to see someone actually get close to what they can do with GPT-3.5/4, except capable of running on commodity GPUs. What's the most impressive open model so far?

Have you tried falcon 40b instruct? Also take into account that chatgpt likely has some preprompt and by talking to falcon or other OS models it's all in your hands. Furthermore, Not many people discuss the significance of proper output sampling. I myself used to just test open source models with the greedy decoding only. Who knows if they wouldn't even beat (not at all)OpenAI with some clever output sampling scheme.

Does anyone have a link / instructions on getting Falcon 40b to install on Apple Silicon?

Apparently "Hugging Face" have some internal swift code that works (but it has not been released). I'm keen to see how it performs on a maxed out Mac Studio (with all that unified memory available).

Re: GPT-4 details leaked?

#98

Post by TeMPOraL »

> The conspiracy theory that the new GPT-4 quality had been deteriorated might be simply because they are letting the oracle model accept lower probability sequences from the speculative decoding model.

In other words: the speculation was likely right, I'll propose a specific mechanism explaining it, but then still insult the people bringing it up and keep gaslighting them.

Re: GPT-4 details leaked?

#100

Post by juliensalinas »

kyledrake wrote: post #65
PostOnce wrote: post #19

"Open" AI, a charity to benefit us all by pushing and publishing the frontier of scientific knowledge. Nevermind, fuckers, actually it's just to take your jobs and make a few VCs richer. We'll keep the science a secret and try to pressure the government into making it illegal for you to compete with us. https://github.com/ggerganov/llama.cpp https://github.com/openlm-research/open_llama https://huggingface.co/TheBlok…

Unfortunately I've found the current OSS models to be vastly inferior to the OpenAI models. Would love to see someone actually get close to what they can do with GPT-3.5/4, except capable of running on commodity GPUs. What's the most impressive open model so far?

LLaMA 30B or 60B can be very impressive when correctly prompted. Deploying the 60B version is a challenge though and you might need to apply 4-bit quantization with something like https://github.com/PanQiWei/AutoGPTQ or https://github.com/qwopqwop200/GPTQ-for-LLaMa . Then you can improve the inference speed by using https://github.com/turboderp/exllama .

If you prefer to use an "instruct" model à la ChatGPT (i.e. that does not need few-shot learning to output good results) you can use something like this: https://huggingface.co/TheBloke/Wizard-Vicuna-30B-Uncensored... The interesting thing with these Uncensored models is that they don't constantly answer that they cannot help you (which is what ChatGPT and GPT-4 are doing more and more).

Post reply on HN